The overall objective of this proposal is to continue the study of hypothalamic-pituitary-adrenal (H-P-A) axis dysregulation in patients with major depressive disorder, endogenous subtype. We have two major goals: 1) To improve our understanding of the natural history and pathophysiologic processes that account for a higher frequency of H-P-A feedback dysregulation in depressed patients; and 2) to translate this understanding into improved laboratory correlates for medical decision-making with depressed patients. To accomplish these, this project will emphasize serial H-P-A measures during different phases of the affective life cycle, and will study H-P-A dysregulation from an expanded perspective using a battery of pre- and post-dexamethasone plasma cortisol, corticosterone, and B-endorphin/B-lipotropin-like-immunoreactivity. These measures are selected to reflect regulation at both adrenal and pituitary loci in the H-P-A axis. Plasma dexamethasone will be measured as a control variable. All qualifying drug-free patients hospitalized in a clinical research unit will be studied serially before, during, and after treatment during a 30-week follow-up. Non-depressed psychiatric patients and normal individuals will be studied as controls. The treatment design emphasizes blind placebo crossover points. Major hypotheses are: 1) a battery of H-P-A measures reflecting both pituitary and adrenal functions will serve as a more sensitive and specific laboratory correlate of depressive pathophysiology than single measures of adrenal output; 2) H-P-A changes are temporarily related to clinical changes and will monitor clinical progress; 3) pre-treatment H-P-A measures may predict response to somatic antidepressant treatment; 4) persistent H-P-A dysregulation despite treatment ("failure to normalize") will identify patients with either a distinctive pathophysiology, incorrect diagnosis, or inadequate treatment, and will be an excellent predictgor of poor response to treatment and high risk of subsequent relapse; and 5) the resolution of H-P-A dysregulation during treatment will indicate when active antidepressant treatments can be discontinued safely. These hypotheses represent a logical extension of our studies on the dexamethasone suppression test (DST) during year 01-08 of this project. The placebo-controlled crossover design is perhaps the only strategy which will clarify current uncertainties about whether the DST or other H-P-A measures have clinical predictive value.